ORIGINAL ARTICLE

Newborn Screening for Severe Primary ImmunodeficiencyDiseases in Sweden—a 2-Year Pilot TREC and KREC ScreeningStudy

Michela Barbaro1,2 & Annika Ohlsson1,3& Stephan Borte4,5 & Susanne Jonsson1

&

Rolf H. Zetterström1,2& Jovanka King4,6,7 & Jacek Winiarski8,9 & Ulrika von Döbeln1,3

&

Lennart Hammarström4

Received: 11 July 2016 /Accepted: 18 October 2016 /Published online: 21 November 2016# The Author(s) 2016. This article is published with open access at Springerlink.com

Abstract Newborn screening for severe primary immunode-ficiencies (PID), characterized by T and/or B cell lymphope-nia, was carried out in a pilot program in the StockholmCounty, Sweden, over a 2-year period, encompassing 58,834children. T cell receptor excision circles (TREC) and kappa-deleting recombination excision circles (KREC) were mea-sured simultaneously using a quantitative PCR-based methodon DNA extracted from dried blood spots (DBS), with beta-actin serving as a quality control for DNA quantity.Diagnostic cutoff levels enabling identification of newbornswith milder and reversible T and/or B cell lymphopenia werealso evaluated. Sixty-four children were recalled for follow-updue to low TREC and/or KREC levels, and three patients withimmunodeficiency (Artemis-SCID, ATM, and an as yet un-

classified Tcell lymphopenia/hypogammaglobulinemia) wereidentified. Of the positive samples, 24 were associated withprematurity. Thirteen children born to mothers treated withimmunosuppressive agents during pregnancy (azathioprine(n = 9), mercaptopurine (n = 1), azathioprine and tacrolimus(n = 3)) showed low KREC levels at birth, which spontane-ously normalized. Twenty-nine newborns had no apparentcause identified for their abnormal results, but normalizedwith time. Children with trisomy 21 (n = 43) showed a lowermedian number of both TREC (104 vs. 174 copies/μL blood)and KREC (45 vs. 100 copies/3.2 mm blood spot), but onlyone, born prematurely, fell below the cutoff level. Two chil-dren diagnosed with DiGeorge syndrome were found to havelow TREC levels, but these were still above the cutoff level.

Michela Barbaro, Annika Ohlsson and Stephan Borte contributed equallyto this work.

Electronic supplementary material The online version of this article(doi:10.1007/s10875-016-0347-5) contains supplementary material,which is available to authorized users.

* Ulrika von Döbelnulrika.vondebeln@karolinska.se

* Lennart Hammarströmlennart.hammarstrom@ki.se

1 Centre for Inherited Metabolic Diseases, Karolinska UniversityHospital Solna, SE-17176 Stockholm, Sweden

2 Department of Molecular Medicine and Surgery, KarolinskaInstitutet, SE-17176 Stockholm, Sweden

3 Department of Medical Biochemistry and Biophysics, Division ofMolecular Metabolism, Karolinska Institutet,SE-17177 Stockholm, Sweden

4 Department of Clinical Immunology, Karolinska University HospitalHuddinge, SE-14186 Stockholm, Sweden

5 ImmunoDeficiencyCenter Leipzig (IDCL) at Hospital St. GeorgLeipzig, Delitzscher Strasse 141, 04129 Leipzig, Germany

6 Department of Immunopathology, SA Pathology, Women’s andChildren’s Hospital Campus, North Adelaide, South Australia 5006,Australia

7 Robinson Research Institute and Discipline of Paediatrics, School ofMedicine, University of Adelaide, North Adelaide, SouthAustralia 5006, Australia

8 Department of Clinical Technology and Intervention, KarolinskaInstitutet, SE-14186 Stockholm, Sweden

9 Department of Pediatrics, Karolinska University Hospital Huddinge,SE-14186 Stockholm, Sweden

J Clin Immunol (2017) 37:51–60DOI 10.1007/s10875-016-0347-5

This is the first large-scale screening study with a simulta-neous detection of both TREC and KREC, allowing identifi-cation of newborns with both T and B cell defects.

Keywords Newborn screening . primary immunodeficiencydiseases . severe combined immunodeficiency . TREC .

KREC

Introduction

The purpose of neonatal screening is the early recognition oftreatable, mostly genetically determined diseases that manifestwith a high rate of morbidity and mortality. Mass screening ofnewborn infants with dried blood spots (DBS) started in theearly 1960s based on a method developed by Guthrie and Susifor diagnosing phenylketonuria [1]. Peripheral blood from aheel stick was blotted onto filter paper (referred to as BGuthriecard^) at 3 to 5 days after birth, dried, and sent by mail tocentralized laboratories for analysis.

The US congress decided in 2006 to include 29 core and 25secondary disorders in its newborn screening program basedon a priority list of disorders. In recent years, a number ofadditional disorders, among them severe combined immuno-deficiency (SCID), have been included in this list, based onthe well-recognized Wilson and Jungner criteria [2]. InSweden, neonatal screening with DBS was initiated in 1965and today, a total of 24 disorders are included in the program.

Genetically determined disorders of immunity are com-monly referred to as primary immunodeficiencies (PID) andwere first recognized clinically over 60 years ago with theidentification of X-linked agammaglobulinemia (XLA) [3].Today, the group of PID includes more than 250 distinct enti-ties, which are divided into Tcell deficiencies, B cell deficien-cies (the predominant group), combined T and B cell deficien-cies, complement defects, and granulocyte defects [4].

Primary immunodeficiencies as a group are not rare dis-eases, but should be considered in all patients presenting withsevere, atypical, or recurrent infections. The overall preva-lence is unknown, but varies greatly (from 1:600 for IgA de-ficiency, 1:20,000 for common variable immunodeficiency,1:50,000 for SCID, and 1:100,000 for XLA) [4] and refer-ences therein [5]. The estimated incidence of severe forms ofPID, which require immediate attention, is variable in differ-ent populations, but would be expected to be in the order of 5–10 per 100,000 live births.

In 2005, Chan and Puck published a seminal paper, describingthe T-cell receptor excision circle (TREC) assay as a tool forlarge-scale screening for SCID and other T cell lymphopenias[6]. T cell receptor genes are normally edited during T cell dif-ferentiation, and the deleted fragments are circularized and do notundergo further replication in dividing cells. Thus, TRECs are amarker of recently formed T cells. The TREC copy number can

be determined using a quantitative PCR-based method usingDNA extracted from routinely collected DBS. The TREC assaywas implemented in the state of Wisconsin in 2008 [7] and itsinclusion in the newborn screening programs in the USA wasrecommended in 2010. Currently, 38 states are screening forSCID using the TREC assay, with the remaining 12 states dueto commence in 2017 (Jeffrey Modell Foundation, http://www.info4pi.org/).More than threemillion newborns have been testedto date [8]. The results show a frequency of SCID of 1/58,000children (overall 1/7300 with clinically relevant forms of T celllymphopenia) and a high survival rate (92 %) after treatment.

Congenital B cell lymphopenia can be identified by screen-ing for kappa-deleting recombination excision circles(KREC), the circular by-product of B cell immunoglobulinkappa gene rearrangement [9]. Patients with severe forms ofB cell deficiency such as XLA are easily identified by theKREC assay [10] and are treated with regular gammaglobulininfusions, thereby allowing a near to normal life. Furthermore,it has been shown that delayed-onset forms of SCID and othersevere forms of PID that present with normal T cell numbersand TREC levels above cutoff in newborn screening programsmay be identified using a combined TREC and KREC screen-ing approach [11].

We thus initiated a screening program for all newborns inthe county of Stockholm in Sweden using a combined TRECand KREC assay [12]. We previously analyzed 10,058 sam-ples as part of a feasibility study, and one XLA patient wasidentified (Bruton tyrosine kinase gene mutation, c.1480C>T,p.Gln494Ter), thus demonstrating proof of principle of thecombined screening approach [12]. Although a small cohortof Spanish newborns have been screened using the combinedTREC/KREC assay [13], this represents the first large-scaleprospective study where both T and B cell defects(lymphopenia) can be identified and the results of the first2 years of the screening program are reported here.

Materials and Methods

Sampling

We carried out a study encompassing samples from all childrenborn in the Stockholm County from November 15, 2013 toNovember 15, 2015, searching for children with T and B celllymphopenia, in order to identify those with severe forms ofPID. A total of 58,834 samples were analyzed. For 24 addition-al infants, the parents chose not to participate in the study. Theblood sample was collected onto Whatman 903 filter paper assoon as possible after 48 h of age, presently at a mean age of2.8 days, and was then mailed to the laboratory. The averageage at recall for a positive screening result was 6 days.

The regional ethical board in Stockholm approved thestudy (Ethical permit 2013/414-31/4).

52 J Clin Immunol (2017) 37:51–60

TREC/KREC Screening Assay

The TREC/KREC newborn screening assay described previ-ously [12] was utilized, with minor technical modifications.Briefly, DNA was extracted from single 3.2-mm punches ofthe original DBS in a 96-well format, and quantitative triplexreal-time qPCR for TREC, KREC, and beta-actin (ACTB)was performed using a ViiA7 Real-Time PCR System(Applied Biosystems, Foster City, CA, USA) on DNA eluatesas previously described [12]. The qPCR procedure was opti-mized using custom reagents provided by Affymetrix (SantaClara, CA, USA). TREC and KREC levels were determinedper 3.2-mmpunch. Amplification of ACTBwas used to assessthat a sufficient quantity of DNA was extracted from eachDBS, but this was disregarded if TREC and KREC copynumbers were above the cutoff level. The samples were ini-tially analyzed once, and if the value was below the cutoff,they were reanalyzed in duplicate. Samples in which TRECand/or KREC levels were below cutoff in association with areduction in ACTB copy number were considered inconclu-sive. Cutoff levels for TREC/KREC were adjusted during thestudy based on the yield of positive samples.

Genetic Characterization of Patients with LowTREC/KREC Levels

For the three children with suspected PID, whole exome se-quencing (WES) was performed, followed by confirmatorySanger sequencing. DNA purification, library preparation,read mapping, variant determination, and analysis of WESdata were performed as described previously [14, 15].Sequences were generated as 90-bp pair-end reads and alignedto the human genome reference (UCSC hg 19 version, build37.1) by using the SOAP aligner (soap 2.21) software.Duplicated reads were filtered out and only uniquely mappedreads were kept for subsequent analysis. The SOAPsnp (ver-sion 1.03) software was subsequently used with default pa-rameters to assemble the consensus sequence and call geno-types in target regions. For single nucleotide polymorphism(SNP) quality control, low-quality SNPs that met one of thefour following criteria were filtered out: (1) a genotype qualityof less than 20, (2) a sequencing depth of less than 4, (3) anestimated copy number of more than 2, and (4) a distance fromthe adjacent SNPs of less than 5 bp. Small insertions/deletionswere detected by using the Unified Genotype tool fromGATK(version v1.0.4705) after alignment of quality reads to thehuman reference genome using BWA (version 0.5.9-r16).

Statistical Analyses

Statistical analyses were performed using GraphPad PrismVersion 6.0 (GraphPad Software, San Diego, CA). Given thenonparametric distribution of the data, the Mann-Whitney U

test was used for group comparison analyses, and theSpearman correlation coefficient was calculated for correla-tion analyses. Differences were considered statistically signif-icant when the p value was less than 0.05.

Results

Screening Overview

In total, 58,834 newborns were screened. Samples with TRECand/or KREC copies below the cutoff values were consideredBabnormal^ (positive), whereas those that also had reducedACTB copy numbers were considered to be Binconclusive.^Initially, 572 samples were found to be either abnormal(positive) or inconclusive (n = 399 and n = 173, respectively)(Fig. 1). After repeat testing on the original DBS, 64 abnormal(positive) samples remained, i.e., 99.9 % were considerednormal and were not evaluated further. A total of 13 patientswere recalled for repeat sampling due to poor quality of theirinitial DBS sample.

The definition of Blow^ levels of TREC and KREC copieswill determine how many samples are considered positive.The dilemma is to avoid unnecessary recalls while still cap-turing all patients with a severe form of PID (i.e., SCID andXLA), as well as additional patients with profound distur-bance of T and B cell homeostasis at birth. Three differentcutoff levels for recall were used during the study as the limitswere adjusted based on the number and outcome of the recalls.During the first period, a cutoff level of 15 copies/3.2 mmblood spot or below for TREC and 10 copies/3.2 mm bloodspot blood or below for KREC was applied. In total, 16,582children were screened and 35 were considered to have apositive result. However, the vast majority recovered sponta-neously (Supplementary Table 1) and the cutoff was subse-quently lowered to a TREC level of 8 copies/3.2 mm bloodspot or below and a KREC level of 4 copies/3.2 mm bloodspot blood or below. Another 28,298 children were screenedduring this period and 11 were found to be positive, amongthem PID patients 1 and 2. As these cutoffs were ultimatelyconsidered too stringent, they were finally changed to a TREClevel of 10 copies/3.2 mm blood spot or below and a KREClevel of 6 copies/3.2 mm blood spot or below. Altogether,13,954 children were subsequently screened and 18 werefound to be positive, among them PID patient 3.

The patients with TREC/KREC levels below the cutoff(Supplementary Table 1) were referred to a pediatrician spe-cialized in the diagnosis and management of PID. The patientwith absent TREC and KREC was immediately hospitalizedwhereas in the vast majority of patients, a follow-up samplefor TREC and KREC analysis was collected within 2–10 weeks (Table 1) depending on the clinical urgency andgestational age. The recall was almost exclusively due to

J Clin Immunol (2017) 37:51–60 53

low KREC levels, and these patients were usually seen andresampled at around 3 weeks of age. In the majority of cases,the TREC/KREC values had normalized at the time of resam-pling (Supplementary Table 1) and only three patients (ap-proximately 1/20,000) were referred for flow cytometric(fluorescence-activated cell sorting, FACS) analysis and ge-netic work-up (WES). In the majority of recalled patients,follow-up investigation with FACS analysis was not per-formed, as it was not deemed clinically relevant. To date, noneof the infants in whom KREC/TREC levels normalized havesubsequently developed features of PID.

Abnormal (Positive) TREC/KREC Results

Altogether, 64 children were recalled for follow-up due to lowTREC and/or KREC levels and three patients with severeprimary immunodeficiency were identified (SupplementaryTable 1). Among the positive samples (Fig. 2), several infantshad more than one potential etiological factor contributing totheir low TREC and/or KREC levels. A total of 24 infantswere premature, including one with trisomy 21 and 11 werea twin or triplet. Furthermore, 13 infants were born to mothersreceiving immunosuppressive therapy. Twenty-nine had noapparent cause identified and the TREC/KREC levels in thechildren tested (n = 27) normalized with time. Resamplingwas declined for four of the 64 children recalled.

The 13 children born to mothers who were treated withimmunosuppressive agents during pregnancy (azathioprine(n = 9), mercaptopurine (n = 1), azathioprine and tacrolimus(n = 3)) showed low KREC levels at birth, which howeverspontaneously normalized after 2–10 weeks. Two childrenfrom a mother treated with azathioprine, born almost 2 yearsapart, both showed low KREC at birth.

Other Conditions Influencing the Screening Result

Children with trisomy 21 (n = 43) showed a lower mediannumber of both TREC (104 vs. 174 in the total material) andKREC (45 vs. 100), but only one child, born prematurely, fellbelow the cutoff levels applied at the time of testing. Twochildren were diagnosed with DiGeorge syndrome in theDepartment of Clinical Genetics, and both had a low TREClevel but these were still above the cutoff level at the time oftesting.

Among the 3457 infants born prematurely (prior to37 weeks gestation), 24 screened positive with low TREC(n = 12) , KREC (n =11) l eve l s , o r bo th (n = 1)(Supplementary Table 1). Although T cells appear to maturelate in fetal life, TREC and KREC levels were surprisinglyBnormal^ in the majority of children, even at very low gesta-tional ages (Table 2), albeit as a group they were still signifi-cantly below the levels in children born at term (p < 0.0001,Mann-Whitney U test) (Fig. 3).

A total of 896 twin pairs and 16 triplet sets (n =1840) wereidentified in the cohort, and the TREC and/or KREC valuesdiffered markedly between the newborns in both those sib-lings with low levels (Table 3) and in those with values withinthe normal range (Fig. 4).

To the best of our knowledge, no patients with severe PIDcharacterized by absent T or B cells at birth have been missedto date among those infants included in the screeningprogram.

PID Patients

Three patients were found to have repeatedly low levels of bothTREC and KREC (n= 1) or TREC alone (n = 2) (Table 1) and

Fig. 1 Summary of all newborn screening results between 15 November 2013 and 15November 2015. *Two infants had low TREC levels, and one hadlow TREC and KREC levels. **See Fig. 2

54 J Clin Immunol (2017) 37:51–60

were considered to have a severe immunodeficiency disorder.The first patient had a homozygous splice-site mutation inArtemis (DLCLRE1C c.333+2T>G) with absent protein expres-sion and underwent a successful stem cell transplantation at theage of 2 months. The second child shows clinical features ofataxia-telangiectasia and carries compound heterozygous muta-tions in ATM, where the mother contributed a c.3673C>T muta-tion resulting in a stop codon (p.Gln1225Ter) and the fathercontributed a c.8653_8654insT mutation resulting in aVal2886CysfsTer10 mutation. The third child was found to haveT cell lymphopenia and hypogammaglobulinemia, but geneticanalysis (including whole genome sequencing) has not as yetrevealed any causativemutation in known PID genes. The resultsof immunological investigations for the three PID patients iden-tified are given in Supplementary Table 2.

Discussion

Since 2008, TREC analysis has been the method of choice forscreening of newborns for severe forms of primary T celllymphopenia [6, 7, 16], a method that will capture most, butnot all, children with SCID (exceptions being patients with

mutations in ZAP70 [17, 18], MHC class II [19–21], andADA (delayed-onset disease [11]). We subsequently devel-oped a triplex method that also includes KREC analysis toassess for potential B cell lymphopenia in DBS samples.This method can successfully identify patients with XLA[12], selected patients with delayed-onset adenosine deami-nase deficiency (ADA) [11], Nijmegen-breakage syndrome(NBS) [12], and purine nucleoside phosphorylase (PNP) de-ficiency [22]. The combined assay has recently also been usedto screen patients with SCID [23, 24], ATM [12, 25], Wiskott-Aldrich syndrome (WAS) [26], DiGeorge syndrome [27–29],and trisomy 21 [30], and it has been suggested to be includedin routine screening of newborns for primary immunodefi-ciency [31–34]. KREC levels have also been used to monitorimmune reconstitution after bone marrow/stem cell transplan-tation [35–37].

Ours is the largest study to date using a combinedTREC/KREC assay for newborn screening. One SCID patientwas identified (and successfully transplanted) in our cohort of58,000 infants, comparable with the data published by Kwanet al. on a very large US cohort [8]. One patient with ATM wasalso identified. This disorder is very rare, with an estimatedincidence of 1 in 200,000 according to recent figures from the

Fig. 2 Characteristics of the 64infants with abnormal screeningresults recalled for repeat testing.*Azathioprine (n = 9),mercaptopurine (n = 1),azathioprine + tacrolimus (n = 3).**One infant who was prematurealso had trisomy 21

Table 1 Serial TREC and KREC levels in PID patients with abnormal screening test results

TREC copies/3.2 mm blood spot KREC copies/3.2 mm blood spot

Case Sex Gestationalage (weeks)

Diagnosis Result 1 Result 2 Result 3 Result 4 Result 1 Result 2 Result 3 Result 4

1 Male 34 Artemis deficiency Day 20

Day 80

Day 140

Day 240

Day 20

Day 80

Day 140

Day 240

2 Male 39 Ataxia-telangiectasia Day 25

Day 255

Day 433

Day 513

Day 27

Day 2510

Day 4321

Day 5119

3 Male 36 Unknown genetic defecta Day 27

Day 274

Day 345

Day 941

Day 2205

Day 27732

Day 34504

Day 94250

a Tcell lymphopenia and hypogammaglobulinemia. At follow-up at 15months of age, the child’s hypogammaglobulinemia had resolved but a persistentidiopathic CD3+ T cell lymphopenia (absolute CD3+ count = 0.9 × 109 /L) was evident

J Clin Immunol (2017) 37:51–60 55

USA [38]. It will be interesting to determine if this finding is dueto serendipity or if the disease is more frequent in Scandinavia.Our third patient showed a T cell lymphopenia of unknowncause. Such cases have been previously described in studies inthe USAwith incidences varying between 3 and 4 per 100,000children in different states and with a predicted incidence of 1 in160,000 in a British study (Professor Bobby Gaspar, GreatOrmond Street Hospital, personal communication). Recently,

during the third year of screening, yet another SCID patient(ADA deficiency) and a patient with severe T cell lymphopeniahave been identified.

The number of referrals for flow cytometric analysis in ourcohort was low (approximately 1:20,000), markedly lower thanthat reported in the US study of Kwan et al. [8] (ranging from1/735 to 1/7500 in the different states). This is a reflection of thestrict resampling strategy that was applied to children with low,

Table 2 Mean TREC and KREClevels at different gestational ages Gestational

age (weeks)Infants (n) Mean TREC level

(copies/3.2 mm blood spot)Mean KREC level(copies/3.2 mm blood spot)

20 1 262 139

21 2 97 180

22 3 84 97

23 23 53 52

24 31 68 58

25 38 64 70

26 50 71 62

27 58 89 65

28 53 93 76

29 97 114 91

30 97 139 110

31 122 134 99

32 199 146 90

33 283 139 87

34 390 142 93

35 663 153 97

36 1347 156 103

Total 3457

< 27 weeks 27 - 36 weeks > 37 weeks

0

500

1000

KR

EC

co

pies/3.2m

m b

lo

od

sp

ot

KREC levels according to gestational age

**** p < 0.0001****

**** ****

n = 148 n = 3309 n = 55377

< 25 weeks 25 - 36 weeks > 37 weeks

0

500

1000

TR

EC

c

op

ie

s/3

.2

mm

b

lo

od

s

po

t

TREC levels according to gestational age

****

**** p < 0.0001**** ****

n = 60 n = 3397 n = 55377

A B

Fig. 3 Comparison of TREC levels (a) in infants born prior to 25 weeks,between 25 and 36 weeks, and at term (≥37 weeks) gestation.Comparison of KREC levels (b) in infants born prior to 27 weeks,between 27 and 36 weeks, and at term (≥37 weeks) gestation. Eachpoint represents one infant and the number of infants in each group is

indicated. The horizontal blue line indicates the mean value of allsamples. The dashed horizontal red line represents the cutoff value forTREC (<10 copies/3.2 mm blood spot) and KREC (<6 copies/3.2 mmblood spot) levels, respectively. ****p < 0.0001, Mann-Whitney U test

56 J Clin Immunol (2017) 37:51–60

but not absent TREC or KREC levels. These results also indicatethat the addition of KREC testing to the TREC assay in a new-born screening program does not critically drive the demand forsecond tier testing nor does it increase the fiscal impact of the testitself due to the minimal additional cost of multiplexing.

Prematurity has been considered a predisposing factor forlow TREC numbers in studies in American infants [8]. TheTREC and KREC levels in our study did show a downwardtrend with decreasing gestational age, although in the vastmajority, the levels were above the cutoff, even in markedly

Table 3 Discrepant TREC andKREC levels in twin and tripletsets with abnormal screeningresults

Gestationalage (weeks)

Sex Twin/triplet

Age at firstsampling(days)

TRECcopies/3.2 mmblood spot

KRECcopies/3.2 mmblood spot

Comments

1aa 38 Female TW I 3 1 18 Deceased

1b 38 Female TW II 3 138 161

2a 34 Male TW I 2 51 259

2ba 34 Male TW II 2 13 92

3aa 25 Female TW I 0 5 25 Deceased

3b 25 Female TW II 0 19 178

4a 29 Male TR I 3 20 42

4ba 29 Female TR II 3 15 29

4c 29 Female TR III 3 70 56

5a 29 Male TW I 2 76 130

5ba 29 Male TW II 2 5 16

6a 27 n.a TR I Deceased

6b 27 n.a TR II Deceased

6ca 27 Male TR III 3 5 28

7aa 38 Female TW I 3 149 6

7ba 38 Female TW II 3 132 5

8a 32 Female TW I 2 77 40

8ba 32 Female TW II 2 13 5

9a 24 Male TW I 2 16 68

9ba 24 Female TW II 2 9 13

10a 23 n.a TW I Deceased

10ba 23 Female TW II 2 7 11

TW twin 1, TR triplet, n.a. not availablea Recalled twin/triplet

0 200 400 600

0

200

400

600

Correlation between TREC levels within each twin pair

TREC copies/3.2mm blood spot

Twin 1

TR

EC

co

pie

s/3

.2m

m b

loo

d s

po

t

Tw

in 2

rs= 0.60

0 100 200 300 400

0

100

200

300

400

Correlation between KREC levels within each twin pair

KREC copies/3.2mm blood spot

Twin 1

KR

EC

co

pies/3

.2m

m b

loo

d s

po

t

Tw

in 2

rs= 0.57

A B

Fig. 4 Correlation between TREC (a) and KREC (b) levels within all896 twin pairs. Each point represents one twin pair, where the TRECvalue for twin 1 (x axis) is plotted against the TREC value for twin 2 (yaxis) (a). KREC levels for each pair are similarly depicted (b). Overall,

there was poor correlation between specimen pairs, with a calculatedSpearman correlation coefficient (Rs) of 0.6 for TREC and 0.57 forKREC (p < 0.0001)

J Clin Immunol (2017) 37:51–60 57

premature infants. This notion was also recently supported[13] in a small Spanish cohort of newborns.

Twins and triplets were markedly overrepresented among thepositive samples (n = 11/64), potentially reflecting prematurity inthemajority of cases (n= 8/11). There was a notable difference inTREC and KREC levels within twin pairs returning normalscreening results. Although information regarding zygosity wasnot available, it is expected that one in four pairs is monozygoticand therefore genetically identical, and this observation will thusbe further investigated in the future studies.

Thirteen children with low KREC levels were born to 12mothers who had been treated with azathioprine during pregnan-cy, three of whomwere concomitantly receiving tacrolimus. Twochildren born to the same azathioprine-treated mother, 2 yearsapart, both had low KREC levels. Low KREC levels were alsorecently noted by de Felipe et al. [13] in a small cohort of Spanishchildren born to mothers treated with azathioprine. As azathio-prine may cross the placenta (and is mutagenic), it is generallycontraindicated during pregnancy unless used for treatment ofsevere disease. As B lymphocytes are markedly more sensitiveto drug-induced apoptosis than T lymphocytes [39], the selectiv-ity for a reduction in KREC copies alone is not surprising.

We anticipated that some infants with trisomy 21 wouldscreen positive since they may have low B cell and/or T cellproduction [30]. However, only one of the 43 infants screenedduring the 2-year period had a KREC level below cutoff,although infants with trisomy 21 had generally lower levelsthan those in the general newborn population. In Sweden, theestimated frequency of chromosome 22q deletions is around1/4000, suggesting that some 14–15 children in the screenedcohort would suffer from this form of PID. However, less than10 % of chromosome 22q deletion patients had pronounced Tcell lymphopenia (measured as low TREC levels) and an ad-ditional 5–10% had levels close to the cutoff. Thus, we wouldhave expected only one to two children to be identified in ourtested cohort, which is not statistically different from the pres-ent result (no child identified).

Most of the children in our study were recalled due to lowKREC levels. This was expected, as we did not knowwhat theBtrue^ cutoff level for children with XLAwould be and thusapplied a fairly large safety margin. Ongoing studies in ourlaboratory, analyzing KREC levels in a large cohort of chil-dren with mutation proven XLA, may allow further adjust-ment of the recall level of KREC, thus reducing the effortsand costs associated with Bfalse^ positive results.

Newborn screening formetabolic diseases has undergonema-jor changes during the past decades with inclusion of numerousvery rare diseases in the program. We anticipate a similar devel-opment in the PID screening field where the KREC may ulti-mately be added to the TREC assay currently used in severalnational screening programs. The decision regarding implemen-tation of TREC vs. combined TREC/KREC screening can beguided by considering the advantages and disadvantages of each

screening approach. The advantages for adding KREC screeninginclude the identification of children with XLA, late onset ADA,some patients with NBS, and other selected forms of PID whosedisease may be undetected by TREC screening alone [8, 23, 40].In addition, it may assist in distinguishing patients with SCIDcaused by deficiency in the production of T as well as B cells oronly T cells, thus aiding in the diagnostic process. On the nega-tive side, increased costs associated with the additional KRECassaymay be considered a disadvantage of the combined screen-ing strategy. With the possibility to multiplex PCR reactions, theindividual cost for additional screeningmarkers such asKREC inan existing TREC screening system is negligible (<€0.10 pernewborn), but associated costs incurred in follow-up and furthertesting for infants with abnormal screening results should also beconsidered. However, it should be noted that the recall rate usingthe combined TREC/KREC assay is similar to that of TRECscreening alone [8], with a lower rate of follow-up FACS studiesthan in other published studies to date.

As markers for other genetic diseases have been proposedwhich can be multiplexed with the TREC/KREC system, fu-ture efforts may be focused on the extension of existing new-born screening tests for PID [26]. This may potentially includecomplement deficiencies (using DBS eluates as described byJanzi et al. [41] for C3 deficiencies and Hamsten et al. [42] forC2 deficiencies) and granulocyte defects, thus allowing cov-erage of a majority of the primary immunodeficiency diseases.Targeted exome sequencing or even whole genome sequenc-ing of newborns may ultimately be considered, which couldlead to identification of additional defects of immunity and alarge number of other inherited, potentially fatal diseases.

Acknowledgments This work was supported by the Jeffrey ModellFoundation and grants provided by the Stockholm county (ALF project20140481) and by an unrestricted research grant from the Baxalta com-pany (BT12-000437). We wish to express our gratitude to KatarinaCarlsson, Jenny Ljusberg-Sjölander, and Fanny Huynh for their excellentlaboratory work.

Authorship Contributions Michela Barbaro, Annika Ohlsson, andSusanne Jonsson performed the screening in the Center for MetabolicDiseases under the supervision of Rolf H. Zetterström and Ulrika vonDöbeln.

Jacek Winiarski is a pediatrician specialized in PID and all childrenwith a suspected PID were referred to him for evaluation and follow-up.

Stephan Borte developed the assay used in this study and was engagedin trouble shooting of the assay during the trial and headed the steeringmeetings during the study.

Jovanka King and Lennart Hammarström wrote the paper.

Compliance with Ethical Standards All procedures performed instudies involving human participants were in accordance with the ethicalstandards of the regional ethical board in Stockholm (Ethical permit2013/414-31/4) and with the 1964 Helsinki declaration and its lateramendments or comparable ethical standards.

Conflict of Interest The authors declare that they have no conflict ofinterest.

58 J Clin Immunol (2017) 37:51–60

Open Access This article is distributed under the terms of the CreativeCommons At t r ibut ion 4 .0 In te rna t ional License (h t tp : / /creativecommons.org/licenses/by/4.0/), which permits unrestricted use,distribution, and reproduction in any medium, provided you giveappropriate credit to the original author(s) and the source, provide a linkto the Creative Commons license, and indicate if changes were made.

References

1. Guthrie R, Susi A. A simple phenylalanine method for detectingphenylketonuria in large populations of newborn infants. Pediatrics.1963;32:338–43.

2. Wilson J, Jungner G. The principles and practice of screening fordisease. Geneva: World Health Organization; 1968.

3. Bruton OC. Agammaglobulinemia. Pediatrics. 1952;9(6):722–8.4. Al-Herz W, Bousfiha A, Casanova JL, Chatila T, Conley ME,

Cunningham-Rundles C, et al. Primary immunodeficiency dis-eases: an update on the classification from the international unionof immunological societies expert committee for primary immuno-deficiency. Front Immunol. 2014;5:162.

5. Bousfiha AA, Jeddane L, Ailal F, Benhsaien I, Mahlaoui N, CasanovaJL, et al. Primary immunodeficiency diseases worldwide: more com-mon than generally thought. J Clin Immunol. 2013;33(1):1–7.

6. Chan K, Puck JM. Development of population-based newbornscreening for severe combined immunodeficiency. J Allergy ClinImmunol. 2005;115(2):391–8.

7. Routes JM, Grossman WJ, Verbsky J, Laessig RH, Hoffman GL,Brokopp CD, et al. Statewide newborn screening for severe T-celllymphopenia. JAMA. 2009;302(22):2465–70.

8. Kwan A, Abraham RS, Currier R, Brower A, Andruszewski K,Abbott JK, et al. Newborn screening for severe combined immu-nodeficiency in 11 screening programs in the United States. JAMA.2014;312(7):729–38.

9. Nakagawa N, Imai K, Kanegane H, Sato H, YamadaM, Kondoh K,et al. Quantification of kappa-deleting recombination excision cir-cles in Guthrie cards for the identification of early B-cell maturationdefects. J Allergy Clin Immunol. 2011;128(1):223–5. e2.

10. Borte S, Wang N, Oskarsdóttir S, von Döbeln U, Hammarström L.Newborn screening for primary immunodeficiencies: beyond SCIDand XLA. Ann N YAcad Sci. 2011;1246:118–30.

11. Speckmann C, Neumann C, Borte S, la Marca G, Sass JO,Wiech E,et al. Delayed-onset adenosine deaminase deficiency: strategies foran early diagnosis. J Allergy Clin Immunol. 2012;130(4):991–4.

12. Borte S, von Döbeln U, Fasth A, Wang N, Janzi M, Winiarski J,et al. Neonatal screening for severe primary immunodeficiency dis-eases using high-throughput triplex real-time PCR. Blood.2012;119(11):2552–5.

13. de Felipe B, Olbrich P, Lucenas JM, Delgado-Pecellin C, Pavon-Delgado A, Marquez J, et al. Prospective neonatal screening forsevere T- and B-lymphocyte deficiencies in Seville. PediatrAllergy Immunol. 2016;27(1):70–7.

14. Alkhairy OK, Perez-Becker R, Driessen GJ, Abolhassani H, vanMontfrans J, Borte S, et al. Novel mutations in TNFRSF7/CD27:clinical, immunologic, and genetic characterization of human CD27deficiency. J Allergy Clin Immunol. 2015;136(3):703–12.e10.

15. Fang M, Abolhassani H, Lim CK, Zhang J, Hammarström L. Nextgeneration sequencing data analysis in primary immunodeficiencydisorders—future directions. J Clin Immunol. 2016.

16. Puck JM, Group SNSW. Population-based newborn screening forsevere combined immunodeficiency: steps toward implementation.J Allergy Clin Immunol. 2007;120(4):760–8.

17. Grazioli S, Bennett M, Hildebrand KJ, Vallance H, Turvey SE,Junker AK. Limitation of TREC-based newborn screening for

ZAP70 severe combined immunodeficiency. Clin Immunol.2014;153(1):209–10.

18. Hauck F, Blumenthal B, Fuchs S, Lenoir C, Martin E, Speckmann C,et al. SYK expression endows human ZAP70-deficient CD8 T cellswith residual TCR signaling. Clin Immunol. 2015;161(2):103–9.

19. Kuo CY, Chase J, Garcia Lloret M, Stiehm ER, Moore T, AguileraMJ, et al. Newborn screening for severe combined immunodefi-ciency does not identify bare lymphocyte syndrome. J AllergyClin Immunol. 2013;131(6):1693–5.

20. Lev A, Simon AJ, Broides A, Levi J, Garty BZ, Rosenthal E, et al.Thymic function in MHC class II-deficient patients. J Allergy ClinImmunol. 2013;131(3):831–9.

21. Lyttle A, Roifman C, Dadi H, Wright N, Kavadas F. MHC class IIdeficiency in the Dene native population: a case report highlightingpitfalls in diagnosis and treatment. Allergy, Asthma Clin Immunol.2014;10(supplement 1):A1.

22. la Marca G, Canessa C, Giocaliere E, Romano F, Malvagia S,Funghini S, et al. Diagnosis of immunodeficiency caused by apurine nucleoside phosphorylase defect by using tandem massspectrometry on dried blood spots. J Allergy Clin Immunol.2014;134(1):155–9.

23. Somech R, Lev A, Simon AJ, Korn D, Garty BZ, Amariglio N,et al. Newborn screening for severe T and B cell immunodeficiencyin Israel: a pilot study. Isr Med Assoc J. 2013;15(8):404–9.

24. Tamura S, Higuchi K, Tamaki M, Inoue C, Awazawa R,Mitsuki N,et al. Novel compound heterozygous DNA ligase IV mutations inan adolescent with a slowly-progressing radiosensitive-severe com-bined immunodeficiency. Clin Immunol. 2015;160(2):255–60.

25. KrausM, LevA, SimonAJ, Levran I, Nissenkorn A, Levi YB, et al.Disturbed B and T cell homeostasis and neogenesis in patients withataxia telangiectasia. J Clin Immunol. 2014;34(5):561–72.

26. Borte S, Meeths M, Liebscher I, Krist K, Nordenskjöld M,Hammarström L, et al. Combined newborn screening for familialhemophagocytic lymphohistiocytosis and severe T- and B-cell im-munodeficiencies. J Allergy Clin Immunol. 2014;134(1):226–8.

27. Lingman Framme J, Borte S, von Döbeln U, Hammarström L,Oskarsdóttir S. Retrospective analysis of TREC based newbornscreening results and clinical phenotypes in infants with the22q11 deletion syndrome. J Clin Immunol. 2014;34(4):514–9.

28. Froňková E, Klocperk A, Svatoň M, Nováková M, Kotrová M,Kayserová J, et al. The TREC/KREC assay for the diagnosis andmonitoring of patients with DiGeorge syndrome. PLoS One.2014;9(12), e114514.

29. Dar N, Gothelf D, Korn D, Frisch A, Weizman A,Michaelovsky E,et al. Thymic and bone marrow output in individuals with 22q11.2deletion syndrome. Pediatr Res. 2015;77(4):579–85.

30. Verstegen RH, Borte S, Bok LA, van Zwieten PH, von Döbeln U,Hammarström L, et al. Impact of Down syndrome on the perfor-mance of neonatal screening assays for severe primary immunode-ficiency diseases. J Allergy Clin Immunol. 2014;133(4):1208–11.

31. Borte S, von Döbeln U, Hammarström L. Guidelines for newbornscreening of primary immunodeficiency diseases. Curr OpinHematol. 2013;20(1):48–54.

32. Chiarini M, Zanotti C, Serana F, Sottini A, Bertoli D, Caimi L, et al.T-cell receptor and K-deleting recombination excision circles innewborn screening of T- and B-cell defects: review of the literatureand future challenges. J Public Health Res. 2013;2(1):9–16.

33. Buelow BJ, Routes JM, Verbsky JW. Newborn screening for SCID:where are we now? Expert Rev Clin Immunol. 2014;10(12):1649–57.

34. Gaspar HB, Hammarström L, Mahlaoui N, Borte M, Borte S. Thecase for mandatory newborn screening for severe combined immu-nodeficiency (SCID). J Clin Immunol. 2014;34(4):393–7.

35. Mensen A, Ochs C, Stroux A, Wittenbecher F, Szyska M, Imberti L,et al. Utilization of TREC and KREC quantification for the monitoringof early T- and B-cell neogenesis in adult patients after allogeneichematopoietic stem cell transplantation. J Transl Med. 2013;11:188.

J Clin Immunol (2017) 37:51–60 59

36. Nakatani K, Imai K, ShigenoM, Sato H, TezukaM, Okawa T, et al.Cord blood transplantation is associated with rapid B-cellneogenesis compared with BM transplantation. Bone MarrowTransplant. 2014;49(9):1155–61.

37. Valotti M, Sottini A, Lanfranchi A, Bolda F, Serana F, Bertoli D, et al.Long-lasting production of new T and B cells and T-cell repertoirediversity in patients with primary immunodeficiency who had under-gone stem cell transplantation: a single-centre experience. J ImmunolRes. 2014;2014:240453.

38. Mallott J, Kwan A, Church J, Gonzalez-Espinosa D, Lorey F,Tang LF, et al. Newborn screening for SCID identifies patientswith ataxia telangiectasia. J Clin Immunol. 2013;33(3):540–9.

39. Dimitriu A, Fauci AS. Activation of human B lymphocytes.XI. Differential effects of azathioprine on B lymphocytes and

lymphocyte subpopulations regulating B cell function. JImmunol. 1978;121(6):2335–9.

40. Chien YH, Chiang SC, Chang KL, Yu HH, Lee WI, Tsai LP, et al.Incidence of severe combined immunodeficiency through newbornscreening in a Chinese population. J Formos Med Assoc.2015;114(1):12–6.

41. Janzi M, Sjöberg R, Wan J, Fischler B, von Döbeln U, Isaac L,et al. Screening for C3 deficiency in newborns using microar-rays. PLoS One. 2009;4(4), e5321.

42. Hamsten C, Skattum L, Truedsson L, von Döbeln U, Uhlén M,Schwenk JM, et al. Heat differentiated complement factor pro-filing. J Proteomics. 2015;126:155–62.

60 J Clin Immunol (2017) 37:51–60